According to Wikipedia, Glicko was invented for games like chess and go. I have also seen references to it in relation to sports. In chess a players skill does not change drastically over a short time. The same would be true for sports where individual or team strength is basically stable over the course of a tournament or season, barring injuries. However using Glicko for WvW does not account for the rapid server strength changes caused by guild migration. I imagine most of us have seen servers gaining or losing strength quickly and then being trapped in either too low or too high a tier. Secondary effects, like leaching Glicko from a rapidly falling server cause addition distortion in the tiers. I would suggest having some group of knowledgeable and impartial people review the Glicko calculations and adjust the matchups where server strength has changed dramatically. A impersonal formula would of course be better, if someone could develop one that could deal with the current realities.
